
import pandas as pd
pd.options.mode.chained_assignment = None
import numpy as np
import matplotlib.pyplot as plt
import re
import seaborn as sns
from textblob import TextBlob
from wordcloud import WordCloud
import plotly.express as px
import string
from nltk.corpus import stopwords
import warnings
warnings.filterwarnings("ignore")
pd.options.display.max_colwidth = 500

C:\Users\Serhan\anaconda3\lib\site-packages\numpy\_distributor_init.py:30: UserWarning: loaded more than 1 DLL from .libs:
C:\Users\Serhan\anaconda3\lib\site-packages\numpy\.libs\libopenblas.PYQHXLVVQ7VESDPUVUADXEVJOBGHJPAY.gfortran-win_amd64.dll
C:\Users\Serhan\anaconda3\lib\site-packages\numpy\.libs\libopenblas.WCDJNK7YVMPZQ2ME2ZZHJJRJ3JIKNDB7.gfortran-win_amd64.dll
  warnings.warn("loaded more than 1 DLL from .libs:\n%s" %


# DATA TAKEN FROM https://www.kaggle.com/gpreda/all-covid19-vaccines-tweets/code
df = pd.read_csv("vaccination_all_tweets.csv", quotechar='"', delimiter=',')
df.head()


print(df.columns)
print(df.shape)

Index(['id', 'user_name', 'user_location', 'user_description', 'user_created',
       'user_followers', 'user_friends', 'user_favourites', 'user_verified',
       'date', 'text', 'hashtags', 'source', 'retweets', 'favorites',
       'is_retweet'],
      dtype='object')
(88978, 16)


# I just need tweets and dates
df=df[['date','text']].dropna()
df['date'] = pd.to_datetime(df['date']).dt.date
df.head()


df = df.drop_duplicates('text')
print(df.shape)

(88860, 2)


# I will delete unwanted content from tweets

texts=df.text
#Removing URLs from tweets
remove_url=lambda x:re.sub(r'http\S+','',str(x))
texts_lr=texts.apply(remove_url)

#Converting all tweets to lowercase
to_lower=lambda x: x.lower()
texts_lr_lc=texts_lr.apply(to_lower)
texts_lr_lc

#Removing punctuations
remove_puncs= lambda x:x.translate(str.maketrans('','',string.punctuation))
texts_lr_lc_np=texts_lr_lc.apply(remove_puncs)
texts_lr_lc_np

#Removing stopwords
more_words=['say','going','like','U','u','#coronavirus', '#coronavirusoutbreak', '#coronavirusPandemic', '#covid19', '#covid_19','coronavirus', 'covid19']
stop_words=set(stopwords.words('english')) #nltk package
stop_words.update(more_words)

remove_words=lambda x: ' '.join([word for word in x.split() if word not in stop_words]) #.join is from package string
texts_lr_lc_np_ns=r=texts_lr_lc_np.apply(remove_words)
texts_lr_lc_np_ns

df.text=texts_lr_lc_np_ns

#Removing emojis
def remove_emoji(text):
    emoji_pattern = re.compile("["
                           u"\U0001F600-\U0001F64F"  # emoticons
                           u"\U0001F300-\U0001F5FF"  # symbols & pictographs
                           u"\U0001F680-\U0001F6FF"  # transport & map symbols
                           u"\U0001F1E0-\U0001F1FF"  # flags (iOS)
                           u"\U00002702-\U000027B0"
                           u"\U000024C2-\U0001F251"
                           "]+", flags=re.UNICODE)
    return emoji_pattern.sub(r'', text)
df['text']=df['text'].apply(lambda x: remove_emoji(x))

display(df.head())


# Sentiment analysis with TextBlob. Polarity ranges from -1 to +1.
df['polarity'] = df['text'].apply(lambda x: TextBlob(x).sentiment.polarity)
df


# Some of the most positive tweets:

df.sort_values(by='polarity', ascending=False)[['text', 'polarity']].reset_index(drop=True).head(n=10)


# Some of the most negative tweets:

df.sort_values(by='polarity', ascending=True)[['text', 'polarity']].reset_index(drop=True).head(n=10)


# Distribution of tweets by polarity

fig = plt.figure(figsize=(10, 6))
df['polarity'].hist()
plt.xlabel('Polarity Score', fontsize=18)
plt.ylabel('Frequency', fontsize=18)
plt.xticks(fontsize=16)
plt.yticks(fontsize=16)

(array([    0., 10000., 20000., 30000., 40000., 50000., 60000., 70000.]),
 [Text(0, 0, ''),
  Text(0, 0, ''),
  Text(0, 0, ''),
  Text(0, 0, ''),
  Text(0, 0, ''),
  Text(0, 0, ''),
  Text(0, 0, ''),
  Text(0, 0, '')])


# I will divide the data into 3 parts for easier processing.

criteria = [df['polarity'].between(-1, -0.01), df['polarity'].between(-0.01, 0.01), df['polarity'].between(0.01, 1)]
values = ['negative', 'neutral', 'positive']
df['sentiment'] = np.select(criteria, values, 0)
fig = plt.figure(figsize=(10, 6))
df['sentiment'].value_counts().sort_index().plot.bar()
plt.xlabel('Sentiment Label', fontsize=18)
plt.ylabel('Tweet Count', fontsize=18)
plt.xticks(fontsize=14)
plt.yticks(fontsize=14)
plt.show()
plt.tight_layout()

<Figure size 432x288 with 0 Axes>


from wordcloud import WordCloud, ImageColorGenerator
import wordninja
from spellchecker import SpellChecker
from collections import Counter
import nltk
import math
import random
from nltk.stem import WordNetLemmatizer
from nltk.corpus import stopwords 
stop_words = set(stopwords.words('english'))  
stop_words.add("amp")


def flatten_list(l):
    return [x for y in l for x in y]

def is_acceptable(word: str):
    return word not in stop_words and len(word) > 2

# Color coding our wordclouds 
def red_color_func(word, font_size, position, orientation, random_state=None,**kwargs):
    return f"hsl(0, 100%, {random.randint(25, 75)}%)" 

def green_color_func(word, font_size, position, orientation, random_state=None,**kwargs):
    return f"hsl({random.randint(90, 150)}, 100%, 30%)" 

def yellow_color_func(word, font_size, position, orientation, random_state=None,**kwargs):
    return f"hsl(42, 100%, {random.randint(25, 50)}%)" 

# Reusable function to generate word clouds 
def generate_word_clouds(neg_doc, neu_doc, pos_doc):
    # Display the generated image:
    fig, axes = plt.subplots(1,3, figsize=(20,10))
    
    wordcloud_neg = WordCloud(max_font_size=50, max_words=100, background_color="white").generate(" ".join(neg_doc))
    axes[0].imshow(wordcloud_neg.recolor(color_func=red_color_func, random_state=3), interpolation='bilinear')
    axes[0].set_title("Negative Words")
    axes[0].axis("off")

    wordcloud_neu = WordCloud(max_font_size=50, max_words=100, background_color="white").generate(" ".join(neu_doc))
    axes[1].imshow(wordcloud_neu.recolor(color_func=yellow_color_func, random_state=3), interpolation='bilinear')
    axes[1].set_title("Neutral Words")
    axes[1].axis("off")

    wordcloud_pos = WordCloud(max_font_size=50, max_words=100, background_color="white").generate(" ".join(pos_doc))
    axes[2].imshow(wordcloud_pos.recolor(color_func=green_color_func, random_state=3), interpolation='bilinear')
    axes[2].set_title("Positive Words")
    axes[2].axis("off")

    plt.tight_layout()
#     plt.show();
    return fig

def get_top_percent_words(doc, percent):
    # Returns a list of "top-n" most frequent words in a list 
    top_n = int(percent * len(set(doc)))
    counter = Counter(doc).most_common(top_n)
    top_n_words = [x[0] for x in counter]
    # print(top_n_words)
    return top_n_words
    
def clean_document(doc):
    spell = SpellChecker()
    lemmatizer = WordNetLemmatizer()
    
    # Lemmatize words (needed for calculating frequencies correctly )
    doc = [lemmatizer.lemmatize(x) for x in doc]
    
    # Get the top 10% of all words. This may include "misspelled" words 
    top_n_words = get_top_percent_words(doc, 0.1)

    # Get a list of misspelled words 
    misspelled = spell.unknown(doc)
    
    # Accept the correctly spelled words and top_n words 
    clean_words = [x for x in doc if x not in misspelled or x in top_n_words]
    
    # Try to split the misspelled words to generate good words (ex. "lifeisstrange" -> ["life", "is", "strange"])
    words_to_split = [x for x in doc if x in misspelled and x not in top_n_words]
    split_words = flatten_list([wordninja.split(x) for x in words_to_split])
    
    # Some splits may be nonsensical, so reject them ("llouis" -> ['ll', 'ou', "is"])
    clean_words.extend(spell.known(split_words))
    
    return clean_words

def get_log_likelihood(doc1, doc2):    
    doc1_counts = Counter(doc1)
    doc1_freq = {
        x: doc1_counts[x]/len(doc1)
        for x in doc1_counts
    }
    
    doc2_counts = Counter(doc2)
    doc2_freq = {
        x: doc2_counts[x]/len(doc2)
        for x in doc2_counts
    }
    
    doc_ratios = {
        # 1 is added to prevent division by 0
        x: math.log((doc1_freq[x] +1 )/(doc2_freq[x]+1))
        for x in doc1_freq if x in doc2_freq
    }
    
    top_ratios = Counter(doc_ratios).most_common()
    top_percent = int(0.1 * len(top_ratios))
    return top_ratios[:top_percent]

# Function to generate a document based on likelihood values for words 
def get_scaled_list(log_list):
    counts = [int(x[1]*100000) for x in log_list]
    words = [x[0] for x in log_list]
    cloud = []
    for i, word in enumerate(words):
        cloud.extend([word]*counts[i])
    # Shuffle to make it more "real"
    random.shuffle(cloud)
    return cloud


def get_smart_clouds(df):

    neg_doc = flatten_list(df[df['sentiment']=='negative']['words'])
    neg_doc = [x for x in neg_doc if is_acceptable(x)]

    pos_doc = flatten_list(df[df['sentiment']=='positive']['words'])
    pos_doc = [x for x in pos_doc if is_acceptable(x)]

    neu_doc = flatten_list(df[df['sentiment']=='neutral']['words'])
    neu_doc = [x for x in neu_doc if is_acceptable(x)]

    # Clean all the documents
    neg_doc_clean = clean_document(neg_doc)
    neu_doc_clean = clean_document(neu_doc)
    pos_doc_clean = clean_document(pos_doc)

    # Combine classes B and C to compare against A (ex. "positive" vs "non-positive")
    top_neg_words = get_log_likelihood(neg_doc_clean, flatten_list([pos_doc_clean, neu_doc_clean]))
    top_neu_words = get_log_likelihood(neu_doc_clean, flatten_list([pos_doc_clean, neg_doc_clean]))
    top_pos_words = get_log_likelihood(pos_doc_clean, flatten_list([neu_doc_clean, neg_doc_clean]))

    # Generate syntetic a corpus using our loglikelihood values 
    neg_doc_final = get_scaled_list(top_neg_words)
    neu_doc_final = get_scaled_list(top_neu_words)
    pos_doc_final = get_scaled_list(top_pos_words)

    # Visualise our synthetic corpus
    fig = generate_word_clouds(neg_doc_final, neu_doc_final, pos_doc_final)
    return fig

# Convert string to a list of words
wordcloud_df = df
wordcloud_df['words'] = wordcloud_df.text.apply(lambda x:re.findall(r'\w+', x ))
get_smart_clouds(wordcloud_df).savefig("sentiment_wordclouds.png", bbox_inches="tight")


# how many tweets by time, let's see visually

timeline = df.groupby(['date']).count().reset_index()
timeline['count'] = timeline['text']
timeline = timeline[['date', 'count']]
fig = px.bar(timeline, x='date', y='count', labels={'date': 'Date', 'count': 'Tweet Count'})
fig.show()


# I'm scaling it to better see how positive/negative tweets change over time
from sklearn.preprocessing import MinMaxScaler

scaler = MinMaxScaler()
df_positive=df[df['sentiment']=='positive'][['date','polarity']].groupby(['date']).mean().reset_index()
df_negative=df[df['sentiment']=='negative'][['date','polarity']].groupby(['date']).mean().reset_index()

df_positive['Positive_scale'] =  scaler.fit_transform(df_positive[["polarity"]])
df_negative['Negative_scale'] =  scaler.fit_transform(df_negative[["polarity"]])
df_positive.drop(columns=['polarity'], inplace=True)
df_negative.drop(columns=['polarity'], inplace=True)


fig = px.bar(df_positive, x='date', y='Positive_scale',title="Positivity Rates of Tweets")
fig.show()
fig = px.bar(df_negative, x='date', y='Negative_scale',title="Negativity  Rates of Tweets")
fig.show()


# DATA TAKEN FROM https://www.kaggle.com/gpreda/covid-world-vaccination-progress
df_vaccine = pd.read_csv("country_vaccinations.csv")
df_vaccine.head()


df_vaccine=df_vaccine[['country','date','total_vaccinations','daily_vaccinations','vaccines']]
df_vaccine.head()


timeline2 = df_vaccine.groupby(['date']).daily_vaccinations.sum()
fig = px.line(timeline2,labels={'date': 'Date', 'value': 'Daily Vaccinations'})
fig.show()


timeline3 = df_vaccine.groupby(['date']).total_vaccinations.sum()
fig = px.line(timeline3,labels={'date': 'Date', 'value': 'Total Vaccinations'})
fig.show()


# Merge polarity of tweets and vaccinations data then scale all

df_daily=timeline2.to_frame()
df_total=timeline3.to_frame()
df_daily.reset_index(drop=False, inplace=True)
df_total.reset_index(drop=False, inplace=True)
df_total['date'] = pd.to_datetime(df_total['date']).dt.date
df_daily['date'] = pd.to_datetime(df_daily['date']).dt.date
df_all=df_daily.merge(df_total)
df_all=df_all.merge(df_positive)
df_all=df_all.merge(df_negative)
df_all['total_vaccinations'] =  scaler.fit_transform(df_all[["total_vaccinations"]])
df_all['daily_vaccinations'] =  scaler.fit_transform(df_all[["daily_vaccinations"]])

df_all


from rdd import rdd


# Regression analysis of Total vaccinations and Positive tweets

model = rdd.rdd(df_all, 'total_vaccinations', 'Positive_scale')
print(model.fit().summary())

Estimation Equation:	 Positive_scale ~ TREATED + total_vaccinations
                            WLS Regression Results                            
==============================================================================
Dep. Variable:         Positive_scale   R-squared:                       0.011
Model:                            WLS   Adj. R-squared:                  0.004
Method:                 Least Squares   F-statistic:                     1.662
Date:                Mon, 07 Jun 2021   Prob (F-statistic):              0.199
Time:                        23:10:09   Log-Likelihood:                 33.099
No. Observations:                 151   AIC:                            -62.20
Df Residuals:                     149   BIC:                            -56.16
Df Model:                           1                                         
Covariance Type:            nonrobust                                         
======================================================================================
                         coef    std err          t      P>|t|      [0.025      0.975]
--------------------------------------------------------------------------------------
Intercept              0.2352      0.010     22.734      0.000       0.215       0.256
TREATED                0.2352      0.010     22.734      0.000       0.215       0.256
total_vaccinations    -0.0755      0.059     -1.289      0.199      -0.191       0.040
==============================================================================
Omnibus:                        4.937   Durbin-Watson:                   1.925
Prob(Omnibus):                  0.085   Jarque-Bera (JB):                4.702
Skew:                           0.431   Prob(JB):                       0.0953
Kurtosis:                       3.070   Cond. No.                     1.78e+15
==============================================================================

Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
[2] The smallest eigenvalue is 9.78e-29. This might indicate that there are
strong multicollinearity problems or that the design matrix is singular.


# Regression analysis of Total vaccinations and Negative tweets

model = rdd.rdd(df_all, 'total_vaccinations', 'Negative_scale')
print(model.fit().summary())

Estimation Equation:	 Negative_scale ~ TREATED + total_vaccinations
                            WLS Regression Results                            
==============================================================================
Dep. Variable:         Negative_scale   R-squared:                       0.037
Model:                            WLS   Adj. R-squared:                  0.031
Method:                 Least Squares   F-statistic:                     5.799
Date:                Mon, 07 Jun 2021   Prob (F-statistic):             0.0173
Time:                        23:10:09   Log-Likelihood:                 73.923
No. Observations:                 151   AIC:                            -143.8
Df Residuals:                     149   BIC:                            -137.8
Df Model:                           1                                         
Covariance Type:            nonrobust                                         
======================================================================================
                         coef    std err          t      P>|t|      [0.025      0.975]
--------------------------------------------------------------------------------------
Intercept              0.3068      0.008     38.858      0.000       0.291       0.322
TREATED                0.3068      0.008     38.858      0.000       0.291       0.322
total_vaccinations     0.1076      0.045      2.408      0.017       0.019       0.196
==============================================================================
Omnibus:                       31.704   Durbin-Watson:                   1.513
Prob(Omnibus):                  0.000   Jarque-Bera (JB):               74.754
Skew:                          -0.866   Prob(JB):                     5.85e-17
Kurtosis:                       5.981   Cond. No.                     1.78e+15
==============================================================================

Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
[2] The smallest eigenvalue is 9.78e-29. This might indicate that there are
strong multicollinearity problems or that the design matrix is singular.


# Regression analysis of Daily vaccinations and Positive tweets

model = rdd.rdd(df_all, 'daily_vaccinations', 'Positive_scale')
print(model.fit().summary())

Estimation Equation:	 Positive_scale ~ TREATED + daily_vaccinations
                            WLS Regression Results                            
==============================================================================
Dep. Variable:         Positive_scale   R-squared:                       0.014
Model:                            WLS   Adj. R-squared:                  0.007
Method:                 Least Squares   F-statistic:                     2.105
Date:                Mon, 07 Jun 2021   Prob (F-statistic):              0.149
Time:                        23:10:10   Log-Likelihood:                 33.321
No. Observations:                 151   AIC:                            -62.64
Df Residuals:                     149   BIC:                            -56.61
Df Model:                           1                                         
Covariance Type:            nonrobust                                         
======================================================================================
                         coef    std err          t      P>|t|      [0.025      0.975]
--------------------------------------------------------------------------------------
Intercept              0.2397      0.012     20.001      0.000       0.216       0.263
TREATED                0.2397      0.012     20.001      0.000       0.216       0.263
daily_vaccinations    -0.0838      0.058     -1.451      0.149      -0.198       0.030
==============================================================================
Omnibus:                        4.304   Durbin-Watson:                   1.930
Prob(Omnibus):                  0.116   Jarque-Bera (JB):                4.052
Skew:                           0.400   Prob(JB):                        0.132
Kurtosis:                       3.060   Cond. No.                     1.77e+15
==============================================================================

Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
[2] The smallest eigenvalue is 1.01e-28. This might indicate that there are
strong multicollinearity problems or that the design matrix is singular.


# Regression analysis of Daily vaccinations and Negative tweets

model = rdd.rdd(df_all, 'daily_vaccinations', 'Negative_scale')
print(model.fit().summary())

Estimation Equation:	 Negative_scale ~ TREATED + daily_vaccinations
                            WLS Regression Results                            
==============================================================================
Dep. Variable:         Negative_scale   R-squared:                       0.041
Model:                            WLS   Adj. R-squared:                  0.035
Method:                 Least Squares   F-statistic:                     6.380
Date:                Mon, 07 Jun 2021   Prob (F-statistic):             0.0126
Time:                        23:10:10   Log-Likelihood:                 74.206
No. Observations:                 151   AIC:                            -144.4
Df Residuals:                     149   BIC:                            -138.4
Df Model:                           1                                         
Covariance Type:            nonrobust                                         
======================================================================================
                         coef    std err          t      P>|t|      [0.025      0.975]
--------------------------------------------------------------------------------------
Intercept              0.3017      0.009     32.998      0.000       0.284       0.320
TREATED                0.3017      0.009     32.998      0.000       0.284       0.320
daily_vaccinations     0.1113      0.044      2.526      0.013       0.024       0.198
==============================================================================
Omnibus:                       31.247   Durbin-Watson:                   1.520
Prob(Omnibus):                  0.000   Jarque-Bera (JB):               74.820
Skew:                          -0.846   Prob(JB):                     5.66e-17
Kurtosis:                       6.005   Cond. No.                     1.77e+15
==============================================================================

Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
[2] The smallest eigenvalue is 1.01e-28. This might indicate that there are
strong multicollinearity problems or that the design matrix is singular.

	id	user_name	user_location	user_description	user_created	user_followers	user_friends	user_favourites	user_verified	date	text	hashtags	source	retweets	favorites	is_retweet
0	1340539111971516416	Rachel Roh	La Crescenta-Montrose, CA	Aggregator of Asian American news; scanning diverse sources 24/7/365. RT's, Follows and 'Likes' will fuel me 👩‍💻	2009-04-08 17:52:46	405	1692	3247	False	2020-12-20 06:06:44	Same folks said daikon paste could treat a cytokine storm #PfizerBioNTech https://t.co/xeHhIMg1kF	['PfizerBioNTech']	Twitter for Android	0	0	False
1	1338158543359250433	Albert Fong	San Francisco, CA	Marketing dude, tech geek, heavy metal & '80s music junkie. Fascinated by meteorology and all things in the cloud. Opinions are my own.	2009-09-21 15:27:30	834	666	178	False	2020-12-13 16:27:13	While the world has been on the wrong side of history this year, hopefully, the biggest vaccination effort we've ev… https://t.co/dlCHrZjkhm	NaN	Twitter Web App	1	1	False
2	1337858199140118533	eli🇱🇹🇪🇺👌	Your Bed	heil, hydra 🖐☺	2020-06-25 23:30:28	10	88	155	False	2020-12-12 20:33:45	#coronavirus #SputnikV #AstraZeneca #PfizerBioNTech #Moderna #Covid_19 Russian vaccine is created to last 2-4 years… https://t.co/ieYlCKBr8P	['coronavirus', 'SputnikV', 'AstraZeneca', 'PfizerBioNTech', 'Moderna', 'Covid_19']	Twitter for Android	0	0	False
3	1337855739918835717	Charles Adler	Vancouver, BC - Canada	Hosting "CharlesAdlerTonight" Global News Radio Network. Weeknights 7 Pacific-10 Eastern - Email comments/ideas to charles@charlesadlertonight.ca	2008-09-10 11:28:53	49165	3933	21853	True	2020-12-12 20:23:59	Facts are immutable, Senator, even when you're not ethically sturdy enough to acknowledge them. (1) You were born i… https://t.co/jqgV18kch4	NaN	Twitter Web App	446	2129	False
4	1337854064604966912	Citizen News Channel	NaN	Citizen News Channel bringing you an alternative news source from citizen journalists that haven't sold out. Real news & real views	2020-04-23 17:58:42	152	580	1473	False	2020-12-12 20:17:19	Explain to me again why we need a vaccine @BorisJohnson @MattHancock #whereareallthesickpeople #PfizerBioNTech… https://t.co/KxbSRoBEHq	['whereareallthesickpeople', 'PfizerBioNTech']	Twitter for iPhone	0	0	False

	date	text
0	2020-12-20	Same folks said daikon paste could treat a cytokine storm #PfizerBioNTech https://t.co/xeHhIMg1kF
1	2020-12-13	While the world has been on the wrong side of history this year, hopefully, the biggest vaccination effort we've ev… https://t.co/dlCHrZjkhm
2	2020-12-12	#coronavirus #SputnikV #AstraZeneca #PfizerBioNTech #Moderna #Covid_19 Russian vaccine is created to last 2-4 years… https://t.co/ieYlCKBr8P
3	2020-12-12	Facts are immutable, Senator, even when you're not ethically sturdy enough to acknowledge them. (1) You were born i… https://t.co/jqgV18kch4
4	2020-12-12	Explain to me again why we need a vaccine @BorisJohnson @MattHancock #whereareallthesickpeople #PfizerBioNTech… https://t.co/KxbSRoBEHq

	date	text
0	2020-12-20	folks said daikon paste could treat cytokine storm pfizerbiontech
1	2020-12-13	world wrong side history year hopefully biggest vaccination effort weve ev…
2	2020-12-12	sputnikv astrazeneca pfizerbiontech moderna russian vaccine created last 24 years…
3	2020-12-12	facts immutable senator even youre ethically sturdy enough acknowledge 1 born i…
4	2020-12-12	explain need vaccine borisjohnson matthancock whereareallthesickpeople pfizerbiontech…

	date	text	polarity
0	2020-12-20	folks said daikon paste could treat cytokine storm pfizerbiontech	0.000
1	2020-12-13	world wrong side history year hopefully biggest vaccination effort weve ev…	-0.500
2	2020-12-12	sputnikv astrazeneca pfizerbiontech moderna russian vaccine created last 24 years…	0.000
3	2020-12-12	facts immutable senator even youre ethically sturdy enough acknowledge 1 born i…	0.100
4	2020-12-12	explain need vaccine borisjohnson matthancock whereareallthesickpeople pfizerbiontech…	0.000
...	...	...	...
88973	2021-05-24	india’s panacea biotec started producing sputnikv russian vaccine first batch produced t…	0.125
88974	2021-05-24	globeandmail canada stop politicizing vaccine toronto ontario ruined lockdown closed business school st…	-0.100
88975	2021-05-24	look arvindkejriwal uselesscm resignnow kejriwalfailsdelhi vaccineshortage sputnikv coronavaccine mumbai…	0.000
88976	2021-05-24	globalnewsto canada stop politicizing vaccine toronto ontario ruined lockdown closed business school st…	-0.100
88977	2021-05-24	rdif panacea biotec start sputnik v production india sputnikv	0.000

	text	polarity
0	vaccinediplomacy gets underway prospects entirely grim sciencediplomacy sputnikv…	-1.0
1	vonderleyen evil face would resemble vaccine virus eu europeanunion…	-1.0
2	awful may rest peace may family get justice deserve vaccine jandj…	-1.0
3	abc7newsbayarea course send faulty vaccines neighborswhy jnj moderna pfizervaccine pathetic po 🤣	-1.0
4	moderna side effects terrible head aches sore arminjection site head aches may bc i…	-1.0
5	indias terrifying indian death toll hits worlds highest…	-1.0
6	day 3 worst headache life moderna covidvaccine	-1.0
7	come eu countries get together save lifes dreadful stop pop at…	-1.0
8	mjlleghari pathetic managed online appointments chughtai muc…	-1.0
9	got secondvaccination yesterday amp feel awful vaccinated covidvaccine moderna	-1.0

INTL 550 - PROJECT¶

SERHAN TÜRKAN¶

SENTIMENT ANALYSIS DURING COVID-19 ¶

DUE TO VACCINATION ROLL-OUTS¶

INTRODUCTION¶

Data Preprocessing¶

Sentiment Analysis¶

Exploratory Data Analysis (EDA)¶

SECOND PART-VACCINATION DATA¶

REGRESSION¶

	text	polarity
0	himantabiswa best part honble narendramodi ji taken jab indias indigenous vaccine covaxin by…	1.0
1	pfizer jab morning efficient wellorganised thank wonderful doctor gav…	1.0
2	pfizer ocugen best vaccine covaxin bharatbiotech	1.0
3	novavax uniofoxford ocugen best vaccine covaxin bharatbiotech	1.0
4	sputnikvaccine americans come russia get sputnikv it’s best vaccine	1.0
5	titanic sinking thank best cm amp vasooli gang lockdownmaharashtra coronavirus…	1.0
6	vaccine best world trust vaccine jai hindcovaxin	1.0
7	lokeshwarri bldatastories annapurani93 businessline srirags best vaccine indian sputnikv …	1.0
8	krysboydthink awesome team moderna	1.0
9	wonderful husband educator became eligible get vaccine asked motivation and…	1.0

	country	iso_code	date	total_vaccinations	people_vaccinated	people_fully_vaccinated	daily_vaccinations_raw	daily_vaccinations	total_vaccinations_per_hundred	people_vaccinated_per_hundred	people_fully_vaccinated_per_hundred	daily_vaccinations_per_million	vaccines	source_name	source_website
0	Afghanistan	AFG	2021-02-22	0.0	0.0	NaN	NaN	NaN	0.0	0.0	NaN	NaN	Oxford/AstraZeneca, Pfizer/BioNTech, Sinopharm/Beijing	World Health Organization	https://covid19.who.int/
1	Afghanistan	AFG	2021-02-23	NaN	NaN	NaN	NaN	1367.0	NaN	NaN	NaN	35.0	Oxford/AstraZeneca, Pfizer/BioNTech, Sinopharm/Beijing	World Health Organization	https://covid19.who.int/
2	Afghanistan	AFG	2021-02-24	NaN	NaN	NaN	NaN	1367.0	NaN	NaN	NaN	35.0	Oxford/AstraZeneca, Pfizer/BioNTech, Sinopharm/Beijing	World Health Organization	https://covid19.who.int/
3	Afghanistan	AFG	2021-02-25	NaN	NaN	NaN	NaN	1367.0	NaN	NaN	NaN	35.0	Oxford/AstraZeneca, Pfizer/BioNTech, Sinopharm/Beijing	World Health Organization	https://covid19.who.int/
4	Afghanistan	AFG	2021-02-26	NaN	NaN	NaN	NaN	1367.0	NaN	NaN	NaN	35.0	Oxford/AstraZeneca, Pfizer/BioNTech, Sinopharm/Beijing	World Health Organization	https://covid19.who.int/

	date	daily_vaccinations	total_vaccinations	Positive_scale	Negative_scale
0	2020-12-12	0.000027	0.000000e+00	0.237758	0.482988
1	2020-12-13	0.000000	5.857430e-07	0.487063	0.664559
2	2020-12-14	0.000022	3.593827e-06	0.168529	0.556242
3	2020-12-15	0.000068	8.166088e-04	0.529504	0.752410
4	2020-12-16	0.006044	1.283449e-05	0.466506	0.524750
...	...	...	...	...	...
146	2021-05-27	0.976800	9.309871e-01	0.261149	0.730174
147	2021-05-28	0.975709	9.239965e-01	0.336346	0.724221
148	2021-05-29	0.987500	9.318942e-01	0.526006	0.725693
149	2021-05-30	1.000000	1.000000e+00	0.339861	0.711905
150	2021-05-31	0.860602	6.990501e-01	0.428659	0.744424